run-13/checkpoint-1500/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.716981132075472,
  "eval_steps": 500,
  "global_step": 1500,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6954838709677419,
      "eval_loss": 0.4076870381832123,
      "eval_runtime": 5.7905,
      "eval_samples_per_second": 535.357,
      "eval_steps_per_second": 11.225,
      "step": 318
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.650593990216632e-05,
      "loss": 0.6428,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8416129032258064,
      "eval_loss": 0.13920199871063232,
      "eval_runtime": 5.9466,
      "eval_samples_per_second": 521.306,
      "eval_steps_per_second": 10.931,
      "step": 636
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8987096774193548,
      "eval_loss": 0.07036732137203217,
      "eval_runtime": 5.6815,
      "eval_samples_per_second": 545.634,
      "eval_steps_per_second": 11.441,
      "step": 954
    },
    {
      "epoch": 3.14,
      "learning_rate": 1.3011879804332637e-05,
      "loss": 0.163,
      "step": 1000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9196774193548387,
      "eval_loss": 0.05059635266661644,
      "eval_runtime": 5.7394,
      "eval_samples_per_second": 540.126,
      "eval_steps_per_second": 11.325,
      "step": 1272
    },
    {
      "epoch": 4.72,
      "learning_rate": 9.517819706498952e-06,
      "loss": 0.0813,
      "step": 1500
    }
  ],
  "logging_steps": 500,
  "max_steps": 2862,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 390310534917408.0,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.5462833281701369,
    "num_train_epochs": 9,
    "temperature": 2
  }
}